US8428303B2 - Postal indicium detection method and postal indicium detection apparatus - Google Patents

Postal indicium detection method and postal indicium detection apparatus Download PDF

Info

Publication number
US8428303B2
US8428303B2 US12/875,943 US87594310A US8428303B2 US 8428303 B2 US8428303 B2 US 8428303B2 US 87594310 A US87594310 A US 87594310A US 8428303 B2 US8428303 B2 US 8428303B2
Authority
US
United States
Prior art keywords
face
postal indicium
detection
picture card
candidate
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related, expires
Application number
US12/875,943
Other versions
US20110091067A1 (en
Inventor
Masaya Maeda
Bunpei Irie
Toshio Sato
Naotake Natori
Tomoyuki Hamamura
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Toshiba Corp
Original Assignee
Toshiba Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Toshiba Corp filed Critical Toshiba Corp
Assigned to KABUSHIKI KAISHA TOSHIBA reassignment KABUSHIKI KAISHA TOSHIBA ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: HAMAMURA, TOMOYUKI, IRIE, BUNPEI, MAEDA, MASAYA, NATORI, NAOTAKE, SATO, TOSHIO
Publication of US20110091067A1 publication Critical patent/US20110091067A1/en
Application granted granted Critical
Publication of US8428303B2 publication Critical patent/US8428303B2/en
Expired - Fee Related legal-status Critical Current
Adjusted expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07B—TICKET-ISSUING APPARATUS; FARE-REGISTERING APPARATUS; FRANKING APPARATUS
    • G07B17/00—Franking apparatus
    • G07B17/04—Franking apparatus with means for avoiding misuse
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
    • B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
    • B07C1/00—Measures preceding sorting according to destination
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/14—Image acquisition
    • G06V30/146—Aligning or centring of the image pick-up or image-field
    • G06V30/1463—Orientation detection or correction, e.g. rotation of multiples of 90 degrees
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/40—Document-oriented image-based pattern recognition
    • G06V30/42—Document-oriented image-based pattern recognition based on the type of document
    • G06V30/424—Postal images, e.g. labels or addresses on parcels or postal envelopes
    • G—PHYSICS
    • G07—CHECKING-DEVICES
    • G07B—TICKET-ISSUING APPARATUS; FARE-REGISTERING APPARATUS; FRANKING APPARATUS
    • G07B17/00—Franking apparatus
    • G07B17/00185—Details internally of apparatus in a franking system, e.g. franking machine at customer or apparatus at post office
    • G07B17/00362—Calculation or computing within apparatus, e.g. calculation of postage value
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
    • B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
    • B07C2301/00—Sorting according to destination
    • B07C2301/0058—Verification for the correct postage amount
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition

Definitions

  • Embodiments described herein relate generally to a postal indicium detection method and postal indicium detection apparatus, which detect postal indicia or stamp from sheets such as mail items.
  • a mail processing apparatus In each postal sorting office, a mail processing apparatus is provided.
  • the mail processing apparatus detects postal indicia such as stamps from mail items, and sorts faces and up-and-down directions of mail items based on the detection results of the postal indicia. For example, many stamps include afterglow components.
  • the mail processing apparatus detects afterglow components from mail items and can detect stamps using the detection results of the afterglow components.
  • the mail processing apparatus registers stamp images in advance, and can detect stamps from mail items based on the registered stamp images.
  • Mail items handled by the aforementioned mail processing often include picture cards. On a specific face of each picture card, a picture, photo, or pattern is printed.
  • the aforementioned mail processing apparatus may often erroneously detect a part of the picture, photo, or pattern on the specific face of a picture card as a non-registered stamp having no afterglow component. Due to such detection errors of postal indicia, the mail processing apparatus cannot often normally sort picture cards. That is, such detection errors cause a sorting performance drop of the mail processing apparatus.
  • a mail processing apparatus which can efficiently recognize picture cards. For example, when a character string (e.g., “Printed in”) unique to a picture card is detected from a mail item, this mail processing apparatus detects this mail item as a picture card.
  • a character string e.g., “Printed in”
  • FIG. 1 is a schematic block diagram showing an example of the arrangement of a mail processing apparatus (postal indicium detection apparatus) according to a first embodiment
  • FIG. 2 is a flowchart for explaining an example of a basic operation of postal indicium detection processing by the mail processing apparatus according to the first embodiment
  • FIG. 3 is a view for explaining examples of areas on first and second faces of a letter to which postal indicium detection is applied;
  • FIG. 4 is a flowchart showing an example of picture card detection processing
  • FIG. 5 shows an example of a postal indicium detection target area decision table
  • FIG. 6 is a view showing examples of areas on a first face (address face) and second face (picture card specific face) of a letter to which postal indicium detection is applied;
  • FIG. 7 is a schematic block diagram showing an example of the arrangement of a mail processing apparatus (postal indicium detection apparatus) according to a second embodiment.
  • FIG. 8 is a flowchart for explaining an example of a basic operation of postal indicium detection processing by the mail processing apparatus according to the second embodiment.
  • a postal indicium detection method includes detecting whether or not one of a first face and a second face of a sheet corresponds to a picture card specific face detecting a postal indicium candidate from the second face based on a first detection result indicating that the first face corresponds to the picture card specific face, and detecting the postal indicium candidate from the first face based on a second detection result indicating that the second face corresponds to the picture card specific face.
  • FIG. 1 is a schematic block diagram showing the arrangement of a mail processing apparatus (postal indicium detection apparatus) according to a first embodiment.
  • the mail processing apparatus includes an image input unit 101 , picture card detection unit 102 , sorting face/direction decision unit 103 , postal indicium detection unit 104 , and recognition result output unit 105 .
  • FIG. 2 is a flowchart for explaining an example of a basic operation of postal indicium detection processing by the mail processing apparatus according to the first embodiment.
  • a plurality of letters (mail items/sheets) are fed in turn to the image input unit 101 .
  • the image input unit 101 captures images of a first face (obverse face) and second face (reverse face) of each letter using a visible light scanner to acquire first image data (grayscale image data) from the first face of the letter and second image data (grayscale image data) also from the second face of the letter (ST 101 ).
  • first image data includes the entire image of the first face of the letter
  • the second image data includes the entire image of the second face of the letter
  • the picture card detection unit 102 executes picture card detection processing for the first and second image data (ST 102 ).
  • FIG. 4 is a flowchart showing an example of the picture card detection processing. For example, if the picture card detection unit 102 can detect a specific character string from the first image data, it determines that the letter corresponds to a picture card (ST 201 ). In this case, the picture card detection unit 102 determines the first face of the letter as an address face, and the second face as a picture card specific face. Likewise, if the picture card detection unit 102 can detect the specific character string from second image data, it determines that the letter corresponds to a picture card (ST 201 ).
  • the picture card detection unit 102 determines the second face of the letter as an address face, and the first face as a picture card specific face. Note that when the specific character string can be detected from the first image data, specific character string detection for the second image data may be skipped, or it may also be applied to the second image data to assure certainty.
  • the picture card specific face is a face on which a picture, photo, or pattern is printed.
  • the specific character string is, for example, “Printed in”.
  • an English expression “Printed in” has been exemplified, but expressions in other languages (for example, French and German) may be used in addition to English.
  • the sorting face/direction decision unit 103 specifies an up-and-down direction of the letter (first face) based on the recognition result of the specific character string (an up-and-down direction of the specific character string) detected from the first image data (ST 202 ).
  • the sorting face/direction decision unit 103 specifies an up-and-down direction of the letter (second face) based on the recognition result of the specific character string (an up-and-down direction of the specific character string) detected from the second image data (ST 202 ).
  • the picture card detection unit 102 If the picture card detection unit 102 cannot specify a picture card based on the specific character string, it specifies a picture card (picture card specific face) based on a degree of image complexity. For example, the picture card detection unit 102 detects a degree of image complexity of the first image data, and detects based on the detection result of the image complexity whether or not the first face corresponding to the first image data corresponds to a picture card specific face. Likewise, the picture card detection unit 102 detects a degree of image complexity of the second image data, and detects based on the detection result of the image complexity whether or not the second face corresponding to the second image data corresponds to a picture card specific face.
  • the picture card detection unit 102 determines that the first face corresponds to a picture card specific face. That is, the picture card detection unit 102 determines that the letter with this first face corresponds to a picture card.
  • the picture card detection unit 102 determines that the second face corresponds to a picture card specific face. That is, the picture card detection unit 102 determines that the letter with this second face corresponds to a picture card.
  • degree of image complexity detection for the second image data may be skipped, or it may also be applied to the second image data to assure certainty.
  • the picture card detection unit 102 can totally detect whether or not the letter corresponds to a picture card or whether or not one of the first and second faces of the letter corresponds to a picture card specific face by combining the aforementioned degree of image complexity detection and specific character string detection. Furthermore, the precision of picture card detection can be improved by utilizing the fact that picture cards have a constant size.
  • the postal indicium detection unit 104 executes normal postal indicium detection processing for the first and second faces (ST 204 ). For example, as shown in FIG. 3 , the postal indicium detection unit 104 attempts postal indicium detection for areas E 1 and E 2 of the first face, and also for areas E 3 and E 4 of the second face, thus detecting a postal indicium (stamp S) from the postal indicium detection attempt results.
  • the sorting face/direction decision unit 103 specifies the up-and-down direction of the second face based on address information which should be recorded on the second face (address face). If the picture card detection unit 102 determines based on the degree of image complexity of the second image data that the second face corresponds to a picture card specific face (YES in ST 203 ), the sorting face/direction decision unit 103 specifies the up-and-down direction of the first face based on address information which should be recorded on the first face (address face).
  • the sorting face/direction decision unit 103 detects, as labels, characters and the like that form address information which should be recorded on the second or first face, and detects continuity given label sizes as first information. Furthermore, the sorting face/direction decision unit 103 detects linearity of the upper or lower ends of characters that form address information which should be recorded on the second or first face as second information. The sorting face/direction decision unit 103 estimates print type rows (address information) from at least one of the first information and the second information (ST 205 ).
  • the sorting face/direction decision unit 103 assumes a first direction of the print type rows as an up direction (normal direction), and executes character recognition of the print type rows in the normal direction.
  • the unit 103 converts the character recognition results into scores, and records a total score (first total score) (ST 206 ).
  • the sorting face/direction decision unit 103 assumes the first direction of the print type rows as a down direction (reverse direction), and executes character recognition of the print type rows in the reverse direction.
  • the unit 103 converts the character recognition results into scores, and records a total score (second total score) (ST 207 ).
  • the sorting face/direction decision unit 103 determines the first direction of the print type rows as the up direction (normal direction) (ST 209 ); if the second total so is equal to or larger than the reference value (YES in ST 210 ), it determines the first direction of the print type rows as the down direction (reverse direction) (ST 211 ). If both the first and second total scores are less than the reference value, the sorting face/direction decision unit 103 determines that a direction cannot be specified (ST 212 ). Note that in this case, the picture card specific face has been specified by the aforementioned processing (ST 212 ).
  • the picture card detection unit 102 detects using at least one of the aforementioned degree of image complexity detection and specific character string detection whether or not the letter corresponds to a picture card or whether or not one of the first and second faces of the letter corresponds to a picture card specific face.
  • the picture card detection unit 102 may detect using other methods whether or not the letter corresponds to a picture card or whether or not one of the first and second faces of the letter corresponds to a picture card specific face.
  • the postal indicium detection unit 104 executes normal postal indicium detection processing for the first and second faces (ST 104 ).
  • the picture card detection unit 102 determines the letter as a picture card (YES in ST 103 ) but the up-end-down direction of the letter cannot be specified (NO in ST 105 ), for example, if the picture card detection unit 102 specifies the letter as a picture card based on the degree of image complexity (YES in ST 203 ), since the up-and-down direction of the letter cannot be specified, the sorting face/direction decision unit 103 specifies the up-and-down direction of the address face based on address information which should be recorded on the address face (ST 106 ).
  • the sorting face/direction decision unit 103 sets the specification results of the up-and-down direction of the address face and letter (ST 107 ), and decides a postal indicium detection target area with reference to a postal indicium detection target area decision table shown in FIG. 5 .
  • a postal indicium detection target area For example, when the first face is specified as an address face, the up-and-down direction of the letter is specified, and the letter is conveyed with the first face facing in the normal direction (up direction), area E 1 shown in FIG. 3 is decided as the postal indicium detection target area.
  • area E 2 shown in FIG. 3 is decided as the postal indicium detection target area.
  • areas E 1 and E 2 shown in FIG. 3 are decided as the postal indicium detection target areas.
  • area E 3 shown in FIG. 3 is decided as the postal indicium detection target area.
  • area E 4 shown in FIG. 3 is decided as the postal indicium detection target area.
  • areas E 3 and E 4 shown in FIG. 3 are decided as the postal indicium detection target areas.
  • areas E 1 and E 3 shown in FIG. 3 are decided as the postal indicium detection target areas.
  • areas E 2 and E 4 shown in FIG. 3 are decided as the postal indicium detection target areas.
  • areas E 1 to E 4 shown in FIG. 3 are decided as the postal indicium detection target areas.
  • the postal indicium detection target area decision table shown in FIG. 5 is provided for an exemplary purpose only, and appropriate decision tables are used according to languages described on letters or the custom of the positions of postal indicia.
  • the postal indicium detection unit 104 detects a postal indicium from the decided detection target area or areas (ST 108 ). For example, the postal indicium detection unit 104 registers postal indicium images in advance, and compares a postal indicium candidate image detected from the decided detection target area with the registered postal indicium images, thereby detecting a postal indicium from the decided detection target area. If a postal indicium cannot be detected (NO in ST 105 ), the postal indicium detection unit 104 outputs postal indicium non-detection information (ST 110 ).
  • the postal indicium detection unit 104 outputs information of a type of the detected postal indicium, a location of the postal indicium, letter faces (positions of an address face and picture card specific dace), and a letter direction.
  • the mail processing apparatus when the mail processing apparatus detects the specific character string from the first face of a letter, it estimates that the second face is a picture card specific face, and executes postal indicium detection processing for the first face without executing it for the second face. In this manner, a picture, photo, or pattern on the picture card specific face can be prevented from being erroneously detected as a postal indicium. For example, as shown in FIG. 6 , even when a picture card specific face includes patterns D 1 and D 2 approximate to actual postal indicia, these patterns D 1 and D 2 can be prevented from being erroneously detected as postal indicia.
  • these patterns D 1 and D 2 are images similar to actual postal indicia, they can be prevented from being erroneously detected as postal indicia.
  • the mail processing apparatus can specify an up-and-down direction of a letter (first face) from the detected specific character string, it can further focus postal indicium detection target areas, thus speeding up the postal indicium detection and improving certainty.
  • the mail processing apparatus detects the specific character string from the second face of a letter, it estimates that the first face is a picture card specific face, and executes postal indicium detection processing for the second face without executing it for the first face. In this manner, a picture, photo, or pattern on the picture card specific face can be prevented from being erroneously detected as a postal indicium.
  • the mail processing apparatus can specify an up-and-down direction of a letter (second face) from the detected specific character string, it can further focus postal indicium detection target areas, thus speeding up the postal indicium detection and improving certainty.
  • the mail processing apparatus When the mail processing apparatus detects the first face as a picture card specific face based on the degree of image complexity, it executes postal indicium detection processing for the second face without executing it for the first face. In this manner, a picture, photo, or pattern on the picture card specific face can be prevented from being erroneously detected as a postal indicium.
  • the mail processing apparatus can specify an up-and-down direction of a letter (second face) from the character recognition result of the second face, it can further focus postal indicium detection target areas, thus speeding up the postal indicium detection and improving certainty.
  • the mail processing apparatus detects the second face as a picture card specific face based on the degree of image complexity, it executes postal indicium detection processing for the first face without executing it for the second face. In this manner, a picture, photo, or pattern on the picture card specific face can be prevented from being erroneously detected as a postal indicium.
  • the mail processing apparatus can specify an up-and-down direction of a letter (first face) from the character recognition result of the first face, it can further focus postal indicium detection target areas, thus speeding up the postal indicium detection and improving certainty.
  • the mail processing apparatus detects a postal indicium from a focused postal indicium detection target area, as described above.
  • a threshold for the postal indicium detection may be dynamically changed.
  • the normal postal indicium detection processing may set a first threshold as that for the postal indicium detection to roughly detect a postal indicium.
  • the postal indicium detection processing using the picture card detection may set a second threshold higher than the first threshold as that for the postal indicium detection so as to severely detect a postal indicium.
  • FIG. 6 assumes a picture card commonly used in English mail items.
  • the mail processing apparatus of this embodiment may assume a picture card commonly used in Japanese mail items to prevent detection errors of postal indicia.
  • detection processing using similarities of images or feature vectors or that of a rectangle or circle by means of Hough transformation may be applied to the postal indicium detection processing.
  • FIG. 7 is a schematic block diagram showing the arrangement of a mail processing apparatus (postal indicium detection apparatus) according to the second embodiment.
  • the mail processing apparatus includes an afterglow image input unit 201 , afterglow detection unit 202 , visible light image input unit 203 , postal indicium detection unit 204 , picture card detection confirmation unit 205 , and recognition result output unit 206 .
  • FIG. 8 is a flowchart for explaining an example of a basic operation of postal indicium detection processing by the mail processing apparatus according to the second embodiment.
  • a plurality of letters (mail items/sheets) are fed in turn the afterglow image input unit 201 .
  • the afterglow image input unit 201 captures images of a first face (obverse face) and second face (reverse face) of each letter using an afterglow scanner to acquire first image data from the first face of the letter and second image data also from the second face of the letter (ST 301 ).
  • the afterglow detection unit 202 detects afterglow components (that of, e.g., a stamp) from the first and second image data (ST 302 ).
  • an appropriate afterglow component is detected as that obtained from, e.g., a stamp.
  • a size of an afterglow component corresponding to that of, e.g., a stamp is registered in advance, and an afterglow component which corresponds to the registered size of the afterglow component (to be referred to as a corresponding afterglow component) is selected from the plurality of afterglow components.
  • noise components included in the corresponding afterglow component are removed.
  • the visible light image input unit 203 captures images of the first face (obverse face) and second face (reverse face) of the letter using a visible light scanner to acquire third image data from the first face of the letter and also fourth image data from the second face of the letter (ST 303 ).
  • the image input processing by the visible light image input unit 203 may be preferentially applied to the letter from which the afterglow component is detected.
  • the postal indicium detection unit 204 then applies postal indicium detection processing to the third image data and also to the fourth image data (ST 304 ). If the postal indicium detection unit 204 cannot detect any postal indicium candidates by the postal indicium detection processing (NO in ST 305 ), it outputs postal indicium non-detection information (ST 306 ). If the postal indicium detection unit 204 can detect one or more postal indicium candidates by the postal indicium detection processing (YES in ST 305 ), and if the detected postal indicium candidates include a postal indicium candidate including the corresponding afterglow component (YES in ST 307 ), the postal indicium candidate including the corresponding afterglow component is selected as a postal indicium (ST 308 ). In this case, since a postal indicium is likely to be correctly detected, the postal indicium detection processing ends without executing postal indicium detection error verification processing (to be described later). In this manner, the high-speed postal indicium detection processing can be maintained.
  • a postal indicium location can be estimated from the detection result of the corresponding afterglow component if no postal indicium candidate is detected.
  • the picture card detection confirmation unit 205 detects whether or not the face of the detected postal indicium candidate corresponds to a sheet specific face (ST 309 ). As described in the first embodiment, the picture card detection confirmation unit 205 detects based on a specific character string detection result whether or not the face of the detected postal indicium candidate corresponds to a sheet specific surface. If a specific character string is detected from the face of the detected postal indicium candidate, it is determined that the face of the detected postal indicium candidate corresponds to a sheet specific surface. Alternatively, as described in the first embodiment, the picture card detection confirmation unit 205 detects based on a degree of image complex not the face of the detected postal indicium candidate corresponds to a sheet specific surface. If the degree of image complexity of the face of the detected postal indicium candidate exceeds a reference value, it is determined that the face of the detected postal indicium candidate corresponds to a picture card specific surface.
  • the recognition result output unit 206 discards the postal indicium candidate included in the picture card specific face (ST 311 ). That is, the postal indicium candidate, which is likely to be erroneously detected, is discarded. In this manner, detection errors of postal indicia can be prevented. Also, as described in the first embodiment, when an up-and-down direction letter is specified, areas which do not include any postal indicium are also discarded according to the contents of the first embodiment (ST 312 ).
  • the recognition result output unit 206 outputs postal indicium non-detection information (ST 314 ). In this way, detection errors of postal indicia can be prevented even when no postal indicium is detected. Detection errors of postal indicia cause sorting error and seal errors of letters. If such errors have occurred, heavy load is imposed on workers. That is, prevention of detection errors of postal indicia mean much even when postal indicia cannot be detected.
  • the recognition result output unit 206 compares scores of postal indicium likelihoods of these plurality of postal indicium candidates, and selects a postal indicium candidate closest to a postal indicium (ST 315 ). Then, the recognition result output unit 206 determines the selected postal indicium candidate as a postal indicium, and outputs information of a type of this postal indicium, a location of the postal indicium, letter faces (positions of an address face and picture card specific face), and letter direction (ST 316 ).
  • the recognition result output unit 206 outputs postal indicium non-detection information.
  • the mail processing apparatus can confirm correctness of the detected postal indicium candidate, and can consequently prevent detection errors of postal indicia.
  • the mail processing apparatus en the postal indicium detection processing without executing the subsequent postal indicium detection error verification processing. In this manner, the high-speed postal indicium detection processing can be maintained.
  • the mail processing apparatus can prevent a part of a picture, photo, or pattern on a specific face of a picture card from being detected as a postal indicium including no afterglow component. Since detection errors of postal indicia are prevented, a sorting performance drop of the mail processing apparatus can be further prevented. Since the mail processing apparatus can focus postal indicium detection target areas by the picture card detection processing, the letter processing speed can be improved.

Abstract

According to one embodiment, a postal indicium detection method includes detecting whether or not one of a first face and a second face of a sheet corresponds to a picture card specific face, detecting a postal indicium candidate from the second face based on a first detection result indicating that the first face corresponds to the picture card specific face, and detecting the postal indicium candidate from the first face based on a second detection result indicating that the second face corresponds to the picture card specific face.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS
This application is based upon and claims the benefit of priority from Japanese Patent Application No. 2009-239505, filed Oct. 16, 2009; the entire contents of which are incorporated herein by reference.
FIELD
Embodiments described herein relate generally to a postal indicium detection method and postal indicium detection apparatus, which detect postal indicia or stamp from sheets such as mail items.
BACKGROUND
In each postal sorting office, a mail processing apparatus is provided. The mail processing apparatus detects postal indicia such as stamps from mail items, and sorts faces and up-and-down directions of mail items based on the detection results of the postal indicia. For example, many stamps include afterglow components. The mail processing apparatus detects afterglow components from mail items and can detect stamps using the detection results of the afterglow components. The mail processing apparatus registers stamp images in advance, and can detect stamps from mail items based on the registered stamp images.
Mail items handled by the aforementioned mail processing often include picture cards. On a specific face of each picture card, a picture, photo, or pattern is printed. The aforementioned mail processing apparatus may often erroneously detect a part of the picture, photo, or pattern on the specific face of a picture card as a non-registered stamp having no afterglow component. Due to such detection errors of postal indicia, the mail processing apparatus cannot often normally sort picture cards. That is, such detection errors cause a sorting performance drop of the mail processing apparatus.
On the other hand, a mail processing apparatus which can efficiently recognize picture cards has been disclosed. For example, when a character string (e.g., “Printed in”) unique to a picture card is detected from a mail item, this mail processing apparatus detects this mail item as a picture card.
Although the mail processing apparatus which can efficiently recognize picture cards has been disclosed, as described above, it is still difficult to prevent the aforementioned detection errors of postal indicia. That is, it is difficult to improve the sorting performance of mail items unless detection errors of postal indicia can be reduced.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a schematic block diagram showing an example of the arrangement of a mail processing apparatus (postal indicium detection apparatus) according to a first embodiment;
FIG. 2 is a flowchart for explaining an example of a basic operation of postal indicium detection processing by the mail processing apparatus according to the first embodiment;
FIG. 3 is a view for explaining examples of areas on first and second faces of a letter to which postal indicium detection is applied;
FIG. 4 is a flowchart showing an example of picture card detection processing;
FIG. 5 shows an example of a postal indicium detection target area decision table;
FIG. 6 is a view showing examples of areas on a first face (address face) and second face (picture card specific face) of a letter to which postal indicium detection is applied;
FIG. 7 is a schematic block diagram showing an example of the arrangement of a mail processing apparatus (postal indicium detection apparatus) according to a second embodiment; and
FIG. 8 is a flowchart for explaining an example of a basic operation of postal indicium detection processing by the mail processing apparatus according to the second embodiment.
 DETAILED DESCRIPTION
In general, according to one embodiment, a postal indicium detection method includes detecting whether or not one of a first face and a second face of a sheet corresponds to a picture card specific face detecting a postal indicium candidate from the second face based on a first detection result indicating that the first face corresponds to the picture card specific face, and detecting the postal indicium candidate from the first face based on a second detection result indicating that the second face corresponds to the picture card specific face.
FIG. 1 is a schematic block diagram showing the arrangement of a mail processing apparatus (postal indicium detection apparatus) according to a first embodiment. As shown in FIG. 1, the mail processing apparatus includes an image input unit 101, picture card detection unit 102, sorting face/direction decision unit 103, postal indicium detection unit 104, and recognition result output unit 105.
FIG. 2 is a flowchart for explaining an example of a basic operation of postal indicium detection processing by the mail processing apparatus according to the first embodiment. A plurality of letters (mail items/sheets) are fed in turn to the image input unit 101. The image input unit 101 captures images of a first face (obverse face) and second face (reverse face) of each letter using a visible light scanner to acquire first image data (grayscale image data) from the first face of the letter and second image data (grayscale image data) also from the second face of the letter (ST101). For example, assume that the first image data includes the entire image of the first face of the letter, and the second image data includes the entire image of the second face of the letter
The picture card detection unit 102 executes picture card detection processing for the first and second image data (ST102). FIG. 4 is a flowchart showing an example of the picture card detection processing. For example, if the picture card detection unit 102 can detect a specific character string from the first image data, it determines that the letter corresponds to a picture card (ST201). In this case, the picture card detection unit 102 determines the first face of the letter as an address face, and the second face as a picture card specific face. Likewise, if the picture card detection unit 102 can detect the specific character string from second image data, it determines that the letter corresponds to a picture card (ST201). In this case, the picture card detection unit 102 determines the second face of the letter as an address face, and the first face as a picture card specific face. Note that when the specific character string can be detected from the first image data, specific character string detection for the second image data may be skipped, or it may also be applied to the second image data to assure certainty.
Note that the picture card specific face is a face on which a picture, photo, or pattern is printed. The specific character string is, for example, “Printed in”. On picture cards (address faces of picture cards), names of places of issuance such as “Printed in Tokyo” and “Printed in France” are often printed. In this case, an English expression “Printed in” has been exemplified, but expressions in other languages (for example, French and German) may be used in addition to English.
As described above, if the picture card detection unit 102 detects the specific character string from the first image data, and determines that the letter corresponds to a picture card (if it determines that the first face corresponds to an address face) (YES in ST201), the sorting face/direction decision unit 103 specifies an up-and-down direction of the letter (first face) based on the recognition result of the specific character string (an up-and-down direction of the specific character string) detected from the first image data (ST202). Likewise, if the picture card detection unit 102 detects the specific character string from the second image data, and determines that the letter corresponds to a picture card (if it determines that the second face corresponds to an address face) (YES in ST201), the sorting face/direction decision unit 103 specifies an up-and-down direction of the letter (second face) based on the recognition result of the specific character string (an up-and-down direction of the specific character string) detected from the second image data (ST202).
If the picture card detection unit 102 cannot specify a picture card based on the specific character string, it specifies a picture card (picture card specific face) based on a degree of image complexity. For example, the picture card detection unit 102 detects a degree of image complexity of the first image data, and detects based on the detection result of the image complexity whether or not the first face corresponding to the first image data corresponds to a picture card specific face. Likewise, the picture card detection unit 102 detects a degree of image complexity of the second image data, and detects based on the detection result of the image complexity whether or not the second face corresponding to the second image data corresponds to a picture card specific face. For example, when the degree of image complexity of the first image data exceeds a reference value, the picture card detection unit 102 determines that the first face corresponds to a picture card specific face. That is, the picture card detection unit 102 determines that the letter with this first face corresponds to a picture card. Likewise, when the degree of image complexity of the second image data exceeds the reference value, the picture card detection unit 102 determines that the second face corresponds to a picture card specific face. That is, the picture card detection unit 102 determines that the letter with this second face corresponds to a picture card. Note that when the degree of image complexity of the first image data exceeds the reference value, degree of image complexity detection for the second image data may be skipped, or it may also be applied to the second image data to assure certainty.
Note that the picture card detection unit 102 can totally detect whether or not the letter corresponds to a picture card or whether or not one of the first and second faces of the letter corresponds to a picture card specific face by combining the aforementioned degree of image complexity detection and specific character string detection. Furthermore, the precision of picture card detection can be improved by utilizing the fact that picture cards have a constant size.
If the picture card detection unit 102 cannot specify a picture card specific face (NO in ST203), the postal indicium detection unit 104 executes normal postal indicium detection processing for the first and second faces (ST204). For example, as shown in FIG. 3, the postal indicium detection unit 104 attempts postal indicium detection for areas E1 and E2 of the first face, and also for areas E3 and E4 of the second face, thus detecting a postal indicium (stamp S) from the postal indicium detection attempt results.
As described above, if the picture card detection unit 102 determines based on the degree of image complexity of the first image data that the first face corresponds to a picture card specific face (YES in ST203), the sorting face/direction decision unit 103 specifies the up-and-down direction of the second face based on address information which should be recorded on the second face (address face). If the picture card detection unit 102 determines based on the degree of image complexity of the second image data that the second face corresponds to a picture card specific face (YES in ST203), the sorting face/direction decision unit 103 specifies the up-and-down direction of the first face based on address information which should be recorded on the first face (address face).
For example, the sorting face/direction decision unit 103 detects, as labels, characters and the like that form address information which should be recorded on the second or first face, and detects continuity given label sizes as first information. Furthermore, the sorting face/direction decision unit 103 detects linearity of the upper or lower ends of characters that form address information which should be recorded on the second or first face as second information. The sorting face/direction decision unit 103 estimates print type rows (address information) from at least one of the first information and the second information (ST205).
Furthermore, the sorting face/direction decision unit 103 assumes a first direction of the print type rows as an up direction (normal direction), and executes character recognition of the print type rows in the normal direction. The unit 103 converts the character recognition results into scores, and records a total score (first total score) (ST206). Moreover, the sorting face/direction decision unit 103 assumes the first direction of the print type rows as a down direction (reverse direction), and executes character recognition of the print type rows in the reverse direction. The unit 103 converts the character recognition results into scores, and records a total score (second total score) (ST207).
If the first total score is equal to or larger than a reference value (YES in ST208), the sorting face/direction decision unit 103 determines the first direction of the print type rows as the up direction (normal direction) (ST209); if the second total so is equal to or larger than the reference value (YES in ST210), it determines the first direction of the print type rows as the down direction (reverse direction) (ST211). If both the first and second total scores are less than the reference value, the sorting face/direction decision unit 103 determines that a direction cannot be specified (ST212). Note that in this case, the picture card specific face has been specified by the aforementioned processing (ST212).
This embodiment will explain a case in which the picture card detection unit 102 detects using at least one of the aforementioned degree of image complexity detection and specific character string detection whether or not the letter corresponds to a picture card or whether or not one of the first and second faces of the letter corresponds to a picture card specific face. However, the picture card detection unit 102 may detect using other methods whether or not the letter corresponds to a picture card or whether or not one of the first and second faces of the letter corresponds to a picture card specific face.
Referring back to the flowchart in FIG. 2, the description will be continued. If the picture card detection unit 102 cannot specify the letter as a picture card (NO in ST103), the postal indicium detection unit 104 executes normal postal indicium detection processing for the first and second faces (ST104). If the picture card detection unit 102 determines the letter as a picture card (YES in ST103) but the up-end-down direction of the letter cannot be specified (NO in ST105), for example, if the picture card detection unit 102 specifies the letter as a picture card based on the degree of image complexity (YES in ST203), since the up-and-down direction of the letter cannot be specified, the sorting face/direction decision unit 103 specifies the up-and-down direction of the address face based on address information which should be recorded on the address face (ST106).
The sorting face/direction decision unit 103 sets the specification results of the up-and-down direction of the address face and letter (ST107), and decides a postal indicium detection target area with reference to a postal indicium detection target area decision table shown in FIG. 5. For example, when the first face is specified as an address face, the up-and-down direction of the letter is specified, and the letter is conveyed with the first face facing in the normal direction (up direction), area E1 shown in FIG. 3 is decided as the postal indicium detection target area. When the first face is specified as an address face, the up-and-down direction of the letter is specified, and the letter is conveyed with the first face facing in the reverse direction (down direction), area E2 shown in FIG. 3 is decided as the postal indicium detection target area. When the first face is specified as an address face, and the up-and-down direction of the letter is not specified, areas E1 and E2 shown in FIG. 3 are decided as the postal indicium detection target areas.
When the second face is specified as an address face, the up-and-down direction of the letter is specified, and the letter is conveyed with the second face facing in the normal direction (up direction), area E3 shown in FIG. 3 is decided as the postal indicium detection target area. When the second face is specified as an address face, the up-and-down direction of the letter is specified, and the letter is conveyed with the second face facing in the reverse direction (down direction), area E4 shown in FIG. 3 is decided as the postal indicium detection target area. When the second face is specified as an address face, and the up-and-down direction of the letter is not specified, areas E3 and E4 shown in FIG. 3 are decided as the postal indicium detection target areas.
When an address face is not specified, the up-and-down direction of the letter is specified, and the letter is conveyed in the normal direction (up direction), areas E1 and E3 shown in FIG. 3 are decided as the postal indicium detection target areas. When an address face is not specified, the up-and-down direction of the letter is specified, and the letter is conveyed in the reverse direction (down direction), areas E2 and E4 shown in FIG. 3 are decided as the postal indicium detection target areas. When an address face is not specified, and the up-and-down direction of the letter is not specified, areas E1 to E4 shown in FIG. 3 are decided as the postal indicium detection target areas.
Note that the postal indicium detection target area decision table shown in FIG. 5 is provided for an exemplary purpose only, and appropriate decision tables are used according to languages described on letters or the custom of the positions of postal indicia.
The postal indicium detection unit 104 detects a postal indicium from the decided detection target area or areas (ST108). For example, the postal indicium detection unit 104 registers postal indicium images in advance, and compares a postal indicium candidate image detected from the decided detection target area with the registered postal indicium images, thereby detecting a postal indicium from the decided detection target area. If a postal indicium cannot be detected (NO in ST105), the postal indicium detection unit 104 outputs postal indicium non-detection information (ST110). If a postal indicium can be detected (YES in ST109), the postal indicium detection unit 104 outputs information of a type of the detected postal indicium, a location of the postal indicium, letter faces (positions of an address face and picture card specific dace), and a letter direction.
As described above, when the mail processing apparatus detects the specific character string from the first face of a letter, it estimates that the second face is a picture card specific face, and executes postal indicium detection processing for the first face without executing it for the second face. In this manner, a picture, photo, or pattern on the picture card specific face can be prevented from being erroneously detected as a postal indicium. For example, as shown in FIG. 6, even when a picture card specific face includes patterns D1 and D2 approximate to actual postal indicia, these patterns D1 and D2 can be prevented from being erroneously detected as postal indicia. Furthermore, even when these patterns D1 and D2 are images similar to actual postal indicia, they can be prevented from being erroneously detected as postal indicia. In addition, since the mail processing apparatus can specify an up-and-down direction of a letter (first face) from the detected specific character string, it can further focus postal indicium detection target areas, thus speeding up the postal indicium detection and improving certainty.
Likewise, when the mail processing apparatus detects the specific character string from the second face of a letter, it estimates that the first face is a picture card specific face, and executes postal indicium detection processing for the second face without executing it for the first face. In this manner, a picture, photo, or pattern on the picture card specific face can be prevented from being erroneously detected as a postal indicium. In addition, since the mail processing apparatus can specify an up-and-down direction of a letter (second face) from the detected specific character string, it can further focus postal indicium detection target areas, thus speeding up the postal indicium detection and improving certainty.
When the mail processing apparatus detects the first face as a picture card specific face based on the degree of image complexity, it executes postal indicium detection processing for the second face without executing it for the first face. In this manner, a picture, photo, or pattern on the picture card specific face can be prevented from being erroneously detected as a postal indicium. In addition, since the mail processing apparatus can specify an up-and-down direction of a letter (second face) from the character recognition result of the second face, it can further focus postal indicium detection target areas, thus speeding up the postal indicium detection and improving certainty.
Likewise, when the mail processing apparatus detects the second face as a picture card specific face based on the degree of image complexity, it executes postal indicium detection processing for the first face without executing it for the second face. In this manner, a picture, photo, or pattern on the picture card specific face can be prevented from being erroneously detected as a postal indicium. In addition, since the mail processing apparatus can specify an up-and-down direction of a letter (first face) from the character recognition result of the first face, it can further focus postal indicium detection target areas, thus speeding up the postal indicium detection and improving certainty.
Note that the mail processing apparatus detects a postal indicium from a focused postal indicium detection target area, as described above. Hence, a threshold for the postal indicium detection may be dynamically changed. For example, the normal postal indicium detection processing may set a first threshold as that for the postal indicium detection to roughly detect a postal indicium. On the other hand, the postal indicium detection processing using the picture card detection may set a second threshold higher than the first threshold as that for the postal indicium detection so as to severely detect a postal indicium.
Note that FIG. 6 assumes a picture card commonly used in English mail items. However, the mail processing apparatus of this embodiment may assume a picture card commonly used in Japanese mail items to prevent detection errors of postal indicia.
Note that detection processing using similarities of images or feature vectors or that of a rectangle or circle by means of Hough transformation may be applied to the postal indicium detection processing.
Mail processing according to the second embodiment will be described below. A description of contents common to the aforementioned first embodiment will not be repeated. FIG. 7 is a schematic block diagram showing the arrangement of a mail processing apparatus (postal indicium detection apparatus) according to the second embodiment. As shown in FIG. 7, the mail processing apparatus includes an afterglow image input unit 201, afterglow detection unit 202, visible light image input unit 203, postal indicium detection unit 204, picture card detection confirmation unit 205, and recognition result output unit 206.
FIG. 8 is a flowchart for explaining an example of a basic operation of postal indicium detection processing by the mail processing apparatus according to the second embodiment. A plurality of letters (mail items/sheets) are fed in turn the afterglow image input unit 201. The afterglow image input unit 201 captures images of a first face (obverse face) and second face (reverse face) of each letter using an afterglow scanner to acquire first image data from the first face of the letter and second image data also from the second face of the letter (ST301). The afterglow detection unit 202 detects afterglow components (that of, e.g., a stamp) from the first and second image data (ST302). For example, after a plurality of afterglow components are detected from the first and second image data by differential labeling processing, an appropriate afterglow component is detected as that obtained from, e.g., a stamp. For example, a size of an afterglow component corresponding to that of, e.g., a stamp is registered in advance, and an afterglow component which corresponds to the registered size of the afterglow component (to be referred to as a corresponding afterglow component) is selected from the plurality of afterglow components. Furthermore, noise components included in the corresponding afterglow component are removed.
Then, the visible light image input unit 203 captures images of the first face (obverse face) and second face (reverse face) of the letter using a visible light scanner to acquire third image data from the first face of the letter and also fourth image data from the second face of the letter (ST303).
Note that the image input processing by the visible light image input unit 203 may be preferentially applied to the letter from which the afterglow component is detected.
The postal indicium detection unit 204 then applies postal indicium detection processing to the third image data and also to the fourth image data (ST304). If the postal indicium detection unit 204 cannot detect any postal indicium candidates by the postal indicium detection processing (NO in ST305), it outputs postal indicium non-detection information (ST306). If the postal indicium detection unit 204 can detect one or more postal indicium candidates by the postal indicium detection processing (YES in ST305), and if the detected postal indicium candidates include a postal indicium candidate including the corresponding afterglow component (YES in ST307), the postal indicium candidate including the corresponding afterglow component is selected as a postal indicium (ST308). In this case, since a postal indicium is likely to be correctly detected, the postal indicium detection processing ends without executing postal indicium detection error verification processing (to be described later). In this manner, the high-speed postal indicium detection processing can be maintained.
Note that when the corresponding afterglow component is detected, a postal indicium location can be estimated from the detection result of the corresponding afterglow component if no postal indicium candidate is detected.
If the detected postal indicium candidates do not include any postal indicium candidate including the corresponding afterglow component (NO in ST307), since a postal indicium is unlikely to be correctly detected, postal indicium detection error verification processing (to be described later) is executed.
The picture card detection confirmation unit 205 detects whether or not the face of the detected postal indicium candidate corresponds to a sheet specific face (ST309). As described in the first embodiment, the picture card detection confirmation unit 205 detects based on a specific character string detection result whether or not the face of the detected postal indicium candidate corresponds to a sheet specific surface. If a specific character string is detected from the face of the detected postal indicium candidate, it is determined that the face of the detected postal indicium candidate corresponds to a sheet specific surface. Alternatively, as described in the first embodiment, the picture card detection confirmation unit 205 detects based on a degree of image complex not the face of the detected postal indicium candidate corresponds to a sheet specific surface. If the degree of image complexity of the face of the detected postal indicium candidate exceeds a reference value, it is determined that the face of the detected postal indicium candidate corresponds to a picture card specific surface.
If the postal indicium candidate is included in a picture card specific face (YES in ST310), the recognition result output unit 206 discards the postal indicium candidate included in the picture card specific face (ST311). That is, the postal indicium candidate, which is likely to be erroneously detected, is discarded. In this manner, detection errors of postal indicia can be prevented. Also, as described in the first embodiment, when an up-and-down direction letter is specified, areas which do not include any postal indicium are also discarded according to the contents of the first embodiment (ST312).
If there is no postal indicium candidate included in a face different from the picture card specific face (NO in ST313), the recognition result output unit 206 outputs postal indicium non-detection information (ST314). In this way, detection errors of postal indicia can be prevented even when no postal indicium is detected. Detection errors of postal indicia cause sorting error and seal errors of letters. If such errors have occurred, heavy load is imposed on workers. That is, prevention of detection errors of postal indicia mean much even when postal indicia cannot be detected.
If there are a plurality of postal indicium candidates included in a face different from the picture card specific face (YES in ST313), the recognition result output unit 206 compares scores of postal indicium likelihoods of these plurality of postal indicium candidates, and selects a postal indicium candidate closest to a postal indicium (ST315). Then, the recognition result output unit 206 determines the selected postal indicium candidate as a postal indicium, and outputs information of a type of this postal indicium, a location of the postal indicium, letter faces (positions of an address face and picture card specific face), and letter direction (ST316).
Note that when the scores of the postal indicium candidates are less than a reference score, the recognition result output unit 206 outputs postal indicium non-detection information.
As described above, the mail processing apparatus can confirm correctness of the detected postal indicium candidate, and can consequently prevent detection errors of postal indicia. In addition, when it is detected that a postal indicium is likely to be correctly detected, the mail processing apparatus en the postal indicium detection processing without executing the subsequent postal indicium detection error verification processing. In this manner, the high-speed postal indicium detection processing can be maintained.
The mail processing apparatus can prevent a part of a picture, photo, or pattern on a specific face of a picture card from being detected as a postal indicium including no afterglow component. Since detection errors of postal indicia are prevented, a sorting performance drop of the mail processing apparatus can be further prevented. Since the mail processing apparatus can focus postal indicium detection target areas by the picture card detection processing, the letter processing speed can be improved.
While certain embodiments have been described, these embodiments have been presented by way of example only, and are not intended to limit the scope of the inventions. Indeed, the novel embodiments described herein may be embodied in a variety of other forms; furthermore, various omissions, substitutions and changes in the form of the embodiments described herein may be made without departing from the spirit of the inventions. The accompanying claims and their equivalents are intended to cover such forms or modifications as would fall within the scope and spirit of the inventions.

Claims (6)

What is claimed is:
1. A postal indicium detection method comprising:
detecting whether or not one of a first face and a second face of a sheet corresponds to a picture card specific face; and
detecting a character direction of the second face based on a first detection result indicating that the first face corresponds to the picture card specific face, estimating a first or second postal indicium detection target area of the second face based on a character direction detection successful result of the second face, and detecting the postal indicium candidate from the first or second postal indicium detection target area of the second face;
detecting the character direction of the second face based on the first detection result, estimating the first and second postal indicium detection target areas of the second face based on a character direction detection error result of the second face, and detecting the postal indicium candidate from the first and second postal indicium detection target areas of the second face;
detecting a character direction of the first face based on the second detection result indicating that the second face corresponds to the picture card specific face, estimating a third or fourth postal indicium detection target area of the first face based on a character direction detection successful result of the first face, and detecting the postal indicium candidate from the third or fourth postal indicium detection target area of the first face; and
detecting the character direction of the first face based on the second detection result, estimating the third and fourth postal indicium detection target areas of the first face based on a character direction detection error result of the first face, and detecting the postal indicium candidate from the third and fourth postal indicium detection target area of the first face.
2. The method of claim 1, further comprising:
detecting based on a degree of complexity of an image of at least one of the first face and the second face whether or not one of the first face and the second face corresponds to the picture card specific face.
3. The method of claim 1, further comprising:
detecting based on a specific character string detection result of at least one of the first face and the second face whether or not one of the first face and the second face corresponds to the picture card specific face.
4. A postal indicium detection method comprising:
detecting, when a postal indicium candidate is detected from a first face of a sheet, whether or not the first face corresponds to a picture card specific face, and determining based on a first detection result indicating that the first face corresponds to the picture card specific face that the postal indicium candidate is likely to be erroneously detected;
detecting a character direction of the second face based on the first detection result, estimating a first or second postal indicium detection target area of the second face based on a character direction detection successful result of the second face, and detecting the postal indicium candidate from the first or second postal indicium detection target area of the second face;
detecting the character direction of the second face based on the first detection result, estimating the first and second postal indicium detection target areas of the second face based on a character direction detection error result of the second face, and detecting the postal indicium candidate from the first and second postal indicium detection target areas of the second face;
detecting, when the postal indicium candidate is detected from a second face of the sheet, whether or not the second face corresponds to the picture card specific face, and determining based on a second detection result indicating that the second face corresponds to the picture card specific face that the postal indicium candidate is likely to be erroneously detected
detecting a character direction of the first face based on the second detection result indicating that the second face corresponds to the picture card specific face, estimating a third or fourth postal indicium detection target area of the first face based on a character direction detection successful result of the first face, and detecting the postal indicium candidate from the third or fourth postal indicium detection target area of the first face; and
detecting the character direction of the first face based on the second detection result, estimating the third and fourth postal indicium detection target areas of the first face based on a character direction detection error result of the first face, and detecting the postal indicium candidate from the third and fourth postal indicium detection target area of the first face.
5. The method of claim 4, further comprising:
detecting, when the postal indicium candidate including no afterglow component is detected from the first face, whether or not the first face corresponds to the picture card specific face, and determining based on the first detection result that the postal indicium candidate is likely to be erroneously detected; and
detecting, when the postal indicium candidate including no afterglow component is detected from the second face, whether or not the second face corresponds to the picture card specific face, and determining based on the second detection result that the postal indicium candidate is likely to be erroneously detected.
6. A postal indicium detection apparatus comprising:
a first detector configured to detect whether or not one of a first face and a second face of a sheet corresponds to a picture card specific face; and
a second detector configured to detect a postal indicium candidate from the second face based on a first detection result indicating that the first face corresponds to the picture card specific face, and to detect the postal indicium candidate from the first face based on a second detection result indicating that the second face corresponds to the picture card specific face,
wherein the second detector detects a character direction of the second face based on a first detection result indicating that the first face corresponds to the picture card specific face, estimates a first or second postal indicium detection target area of the second face based on a character direction detection successful result of the second face, and detects the postal indicium candidate from the first or second postal indicium detection target area of the second face;
the second detector detects the character direction of the second face based on the first detection result, estimates the first and second postal indicium detection target areas of the second face based on a character direction detection error result of the second face, and detects the postal indicium candidate from the first and second postal indicium detection target areas of the second face;
the second detector detects a character direction of the first face based on the second detection result indicating that the second face corresponds to the picture card specific face, estimates a third or fourth postal indicium detection target area of the first face based on a character direction detection successful result of the first face, and detects the postal indicium candidate from the third or fourth postal indicium detection target area of the first face; and
the second detector detects the character direction of the first face based on the second detection result, estimates the third and fourth postal indicium detection target areas of the first face based on a character direction detection error result of the first face, and detects the postal indicium candidate from the third and fourth postal indicium detection target area of the first face.
US12/875,943 2009-10-16 2010-09-03 Postal indicium detection method and postal indicium detection apparatus Expired - Fee Related US8428303B2 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2009-239505 2009-10-16
JP2009239505A JP2011086178A (en) 2009-10-16 2009-10-16 Postal indicium detection method and postal indicium detection apparatus

Publications (2)

Publication Number Publication Date
US20110091067A1 US20110091067A1 (en) 2011-04-21
US8428303B2 true US8428303B2 (en) 2013-04-23

Family

ID=43499745

Family Applications (1)

Application Number Title Priority Date Filing Date
US12/875,943 Expired - Fee Related US8428303B2 (en) 2009-10-16 2010-09-03 Postal indicium detection method and postal indicium detection apparatus

Country Status (5)

Country Link
US (1) US8428303B2 (en)
EP (1) EP2312496A3 (en)
JP (1) JP2011086178A (en)
KR (1) KR101152538B1 (en)
CN (1) CN102043955A (en)

Citations (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB902902A (en) 1957-04-09 1962-08-09 Standard Telephones Cables Ltd Article sorting control apparatus
US4783825A (en) * 1985-04-30 1988-11-08 Kabushiki Kaisha Toshiba Postal material reading apparatus
US5280531A (en) * 1991-10-28 1994-01-18 Pitney Bowes Inc. Apparatus for the analysis of postage meter usage
US5535127A (en) * 1993-07-16 1996-07-09 Kabushiki Kaisha Toshiba Processing apparatus for mail with stamps
JPH08180133A (en) 1994-12-26 1996-07-12 Sharp Corp Image processor
JP2000000524A (en) 1998-06-15 2000-01-07 Nec Corp Method for identifying picture post card and picture postcard identification device
US20010021261A1 (en) * 1998-02-10 2001-09-13 Masashi Koga Address reader, sorting machine such as a mail thing and character string recognition method
JP2004005051A (en) 2002-05-30 2004-01-08 Toshiba Corp Stamp detector and stamp detection method
EP1569178A2 (en) 2004-01-28 2005-08-31 Nec Corporation Image inputting device
US20060044341A1 (en) * 2004-08-30 2006-03-02 Pitney Bowes Incorporated Fluorescent ink detector
JP2007222746A (en) 2006-02-22 2007-09-06 Toshiba Corp Mail processing apparatus and mail processing method
US20080008376A1 (en) 2006-07-07 2008-01-10 Lockheed Martin Corporation Detection and identification of postal indicia
US7668784B2 (en) * 2000-04-05 2010-02-23 Neopost Limited Printing of postal indicia and detection thereof
US7920296B2 (en) * 2007-12-14 2011-04-05 Eastman Kodak Company Automatic determining image and non-image sides of scanned hardcopy media
US8131019B2 (en) * 2007-10-10 2012-03-06 Pitney Bowes Inc. Method and system for capturing images moving at high speed

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS57113877A (en) * 1981-01-07 1982-07-15 Tokyo Shibaura Electric Co Detector for surface on which charge is printed
JPS61275986A (en) * 1985-05-31 1986-12-06 Toshiba Corp Front surface and rear surface deciding device for paper
JPS61276080A (en) * 1985-05-31 1986-12-06 Toshiba Corp Top and reverse surface deciding device
JPH01209247A (en) * 1988-02-12 1989-08-23 Nec Corp Automatic assorter for mail
KR100586312B1 (en) 2004-08-31 2006-06-08 주식회사 퓨트로 Method and system for classifying the postal packages

Patent Citations (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB902902A (en) 1957-04-09 1962-08-09 Standard Telephones Cables Ltd Article sorting control apparatus
US4783825A (en) * 1985-04-30 1988-11-08 Kabushiki Kaisha Toshiba Postal material reading apparatus
US5280531A (en) * 1991-10-28 1994-01-18 Pitney Bowes Inc. Apparatus for the analysis of postage meter usage
US5535127A (en) * 1993-07-16 1996-07-09 Kabushiki Kaisha Toshiba Processing apparatus for mail with stamps
JPH08180133A (en) 1994-12-26 1996-07-12 Sharp Corp Image processor
US20010021261A1 (en) * 1998-02-10 2001-09-13 Masashi Koga Address reader, sorting machine such as a mail thing and character string recognition method
JP2000000524A (en) 1998-06-15 2000-01-07 Nec Corp Method for identifying picture post card and picture postcard identification device
US7668784B2 (en) * 2000-04-05 2010-02-23 Neopost Limited Printing of postal indicia and detection thereof
JP2004005051A (en) 2002-05-30 2004-01-08 Toshiba Corp Stamp detector and stamp detection method
EP1569178A2 (en) 2004-01-28 2005-08-31 Nec Corporation Image inputting device
US20060044341A1 (en) * 2004-08-30 2006-03-02 Pitney Bowes Incorporated Fluorescent ink detector
JP2007222746A (en) 2006-02-22 2007-09-06 Toshiba Corp Mail processing apparatus and mail processing method
US20080008376A1 (en) 2006-07-07 2008-01-10 Lockheed Martin Corporation Detection and identification of postal indicia
US8131019B2 (en) * 2007-10-10 2012-03-06 Pitney Bowes Inc. Method and system for capturing images moving at high speed
US7920296B2 (en) * 2007-12-14 2011-04-05 Eastman Kodak Company Automatic determining image and non-image sides of scanned hardcopy media

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
European Search Report dated Jul. 24, 2012.
Korean Office Action dated Mar. 5, 2012.
NPL-US Postal Postcard specification. *
NPL—US Postal Postcard specification. *

Also Published As

Publication number Publication date
CN102043955A (en) 2011-05-04
EP2312496A2 (en) 2011-04-20
KR101152538B1 (en) 2012-06-01
JP2011086178A (en) 2011-04-28
KR20110041982A (en) 2011-04-22
US20110091067A1 (en) 2011-04-21
EP2312496A3 (en) 2012-08-22

Similar Documents

Publication Publication Date Title
ES2271645T5 (en) LABELING OF IDENTIFICATION OR TIMING OF POST OBJECTS BY IMAGE SIGNATURE AND ASSOCIATED POSTAL CLASSIFICATION MACHINE.
JP5214330B2 (en) Image processing apparatus, barcode processing apparatus, and image processing method
Aradhye A generic method for determining up/down orientation of text in roman and non-roman scripts
JP6286866B2 (en) Image processing apparatus and image processing method
JP2004500974A (en) Automatic location of address information on parcels sent by mass mailers
US8201742B2 (en) Barcode processing apparatus and barcode processing method
JP4855698B2 (en) Address recognition device
JP2000293626A (en) Method and device for recognizing character and storage medium
JP4893278B2 (en) Certificate identification system and certificate identification method
JP4409702B2 (en) Recognition device
US8428303B2 (en) Postal indicium detection method and postal indicium detection apparatus
JP5041775B2 (en) Character cutting method and character recognition device
Wolf et al. Fast address block location on handwritten and machine printed mail-piece images
JP5178851B2 (en) Address recognition device
US20120243728A1 (en) Bulk Region of Interest Learning
US10997452B2 (en) Information processing apparatus and non-transitory computer readable medium storing program
JP2000210624A (en) Mail address recognition device
JP3162552B2 (en) Mail address recognition device and address recognition method
JP2012084132A (en) Character recognition device, character recognition method, and sorting device
JPH0793466A (en) Device for discriminating character kind and method therefor
JP3160347B2 (en) Mail address reading device
JP2015155077A (en) Paper sorting device
US20180004995A1 (en) Systems and methods for detecting double-feed items on mail processing equipment
Tsuchiya et al. A method for determining address format in the automated sorting of Japanese mail
JP2007222746A (en) Mail processing apparatus and mail processing method

Legal Events

Date Code Title Description
AS Assignment

Owner name: KABUSHIKI KAISHA TOSHIBA, JAPAN

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:MAEDA, MASAYA;IRIE, BUNPEI;SATO, TOSHIO;AND OTHERS;REEL/FRAME:024940/0288

Effective date: 20100825

FEPP Fee payment procedure

Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

REMI Maintenance fee reminder mailed
LAPS Lapse for failure to pay maintenance fees
STCH Information on status: patent discontinuation

Free format text: PATENT EXPIRED DUE TO NONPAYMENT OF MAINTENANCE FEES UNDER 37 CFR 1.362

FP Lapsed due to failure to pay maintenance fee

Effective date: 20170423